Teachers' beliefs about students, and the intention of students to drop out of secondary education in Flanders

Teaching and Teacher Education 54 (2016) 117e127

Contents lists available at ScienceDirect

Teaching and Teacher Education journal homepage: www.elsevier.com/locate/tate

Teachers' beliefs about students, and the intention of students to drop out of secondary education in Flanders Mieke Van Houtte*, Jannick Demanet Ghent University, Department of Sociology, CuDOS, Korte Meer 5, 9000 Gent, Belgium

h i g h l i g h t s
We examined the teachers' role in students' intention to dropout from high school.
We related teachers' beliefs about students to students' intentions to quit school.
Being enrolled in vocational education heightened the risk of planning to quit.
Teachers' beliefs impacted dropout plans, beyond students' perceived teacher support.
Particularly in vocational education, teachers' beliefs played an important role.

a r t i c l e i n f o

a b s t r a c t

Article history: Received 13 April 2015 Received in revised form 3 December 2015 Accepted 14 December 2015 Available online xxx

Students dropping out of education with no qualifications has been an important problem in the West for decades. Little relevant research has concentrated on school characteristics, and research into the role teachers might play in students' decision to quit school is particularly scarce. Using multilevel analyses of data for 11,844 students in 84 Flemish secondary schools, we investigate whether teachers' shared expectations of students are associated with students' intention to drop out. Particularly in vocational education, teachers' beliefs about the teachability of students influence the students' intention to quit, irrespective of perceived teacher support and students' sense of futility. © 2015 Elsevier Ltd. All rights reserved.

Keywords: Dropout Teacher beliefs Vocational education Teacher expectations

1. Introduction Students dropping out of education with no qualifications has been an important problem in the West for decades (Lamb, Markussen, Teese, Sandberg, & Polesel, 2011). In Flanders, on average 10%e15% (even higher in the cities) of students in secondary education leave school prematurely, and without any educational qualifications (Van Landeghem, De Fraine, Gielen, & Van Damme, 2013). Predictors of student attrition have been studied intensively, but most existing research has focused on individual student characteristics, specifically sociodemographic and academic risk factors (De Witte, Cabus, Thyssen, Groot, & Maassen van den Brink, 2013), turning dropout into an individual problem (Luyten, Bosker, Dekkers, & Derks, 2003). Relatively little empirical

* Corresponding author. E-mail addresses: [email protected] (M. Van Houtte), Jannick. [email protected] (J. Demanet). http://dx.doi.org/10.1016/j.tate.2015.12.003 0742-051X/© 2015 Elsevier Ltd. All rights reserved.

research has concentrated on school characteristics (Luyten et al., 2003), and research into the role teachers might play in students' decision to quit school is particularly scarce. However, some studies have associated dropout with students' own reports on the quality of teachers (Rumberger & Thomas, 2000), on teachers' relationships with students, and on students' perceived teacher support (Croninger & Lee, 2001; Lee & Burkam, 2003). Perceptions might, however, be shaped after students have left school (see Worrell & Hale, 2001). Moreover, negative feelings about school might bias students' views about their teachers, and accordingly might not inform us about the actual role of teachers in the dropout process (Van Houtte, 2011). Therefore, an assessment that is not obtained from students, but reported by teachers themselvesefor example teachers' expectations or beliefs concerning their studentsemight provide a more accurate picture of the impact of teachers. On a methodological note, research also indicates that surveying a single respondent (teacher or student) when measuring different concepts, might create problems of shared-method variance. Using the

118

M. Van Houtte, J. Demanet / Teaching and Teacher Education 54 (2016) 117e127

same method to measure two variablesein this case, the same individual is surveyedecan yield inflations of intercorrelations and effect sizes, creating bias in the results (Hawker & Boulton, 2000). Accordingly, whereas previous research relating teacher and student characteristics has focused exclusively on either teacher or student data (e.g., Baker, Grant, & Morlock, 2008; Blue & Cook, 2004; Hallinan, 2008), in the current study we relate teacherreported data to student-reported data. More specifically, we examine the relationship between teachers' self-reported beliefs, and students' self-reported intention to drop out. This is made possible by considering individual teacher beliefs as manifestations of the teacher culture at school. These beliefs can be related to student characteristics through a multilevel framework, in which the teacher culture is added as a school-level feature, and the student outcome as an individual-level feature (see also Demanet & Van Houtte, 2012). Consequently, although we lean on the expectancy effects tradition begun by Rosenthal and Jacobson (1968), we do not focus on the expectations individual teachers have about individual students, or on teachers' predictions for particular students (cf. Dalton, Glennie, Ingels, & Wirt, 2009). Instead, we focus on the effect of the teacher culture, in other words, the beliefs teachers in the same school share about their students in general (see alsoVan Houtte, 2004, 2011). To our knowledge, the possible direct influence of teachers' beliefs has rarely been considered in dropout research (for an exception, see Rumberger & Palardy, 2005). However, as teacher beliefs affect both studenteteacher relationships and student engagement (Demanet & Van Houtte, 2012)ethe latter being considered the primary factor in understanding and predicting high school dropout (Fall & Roberts, 2012; Finn, 1989)eit is not too far-fetched to hypothesize an effect of teacher beliefs on students' intention to drop out. Hence, our objective is to investigate how teachers' shared beliefs about studentsethe teacher cultureeaffect the intention of students to drop out of school. In addition, we focus particularly on students in vocational tracks, as the prevalence of dropout is highest in vocational education, making these students most at risk (for Flanders: Van Landeghem & Van Damme, 2011; for the Netherlands: van Uden, Ritzen, & Pieters, 2014; for Turkey: Tas, Selvitopu, Bora, & Demirkaya, 2013). Therefore, knowing that teachers in lower tracks have a relatively poor image of their students (e.g., Kelly & Carbonaro, 2012) and knowing that lower track students, in particular vocational students, are most vulnerable to dropping out, we aim to examine what role the beliefs of teachers play as a determinant of vocational students' intention to drop out. 2. Background 2.1. Antecedents of dropping out Dropping outeleaving secondary education prior to completion, and without qualification or diploma, or with only a minimal credentialehas been studied for decades as a serious educational and social problem (e.g., Blue & Cook, 2004; De Witte et al., 2013; Lee & Burkam, 2003; Rumberger, 1987). By leaving school before completion, most dropouts suffer a high personal cost because of educational deficiencies that hamper their economic and social well-being in adulthood (Rumberger, 1987). There is also a social impact, in terms of the loss of human capital (Lee & Burkam, 2003) and the hindering of policymakers' objectives for sustainable economic growth (De Witte et al., 2013). Consequently, a great deal of research has concentrated on explaining why students leave school unqualified. The most common explanations for dropping out focus on individual students and their families (overview: De Witte et al., 2013; Lee & Burkam, 2003), distinguishing broadly between two

risk factors: social risk and academic risk (Croninger & Lee, 2001; Lee & Burkam, 2003). Social risk entails demographic characteristics associated with a higher likelihood of school failure, such as low socioeconomic status, being male, and having an immigrant background (Blue & Cook, 2004; Dalton et al., 2009; Rumberger, 1987, 1995). Academic risk refers to the actual manifestation of schoolrelated problems, such as poor performance, absenteeism, and grade retention (Blue & Cook, 2004; Croninger & Lee, 2001). In the mid-1990s, Rumberger (1995) started to associate dropout with school features by means of multilevel modeling, dealing with school demographic composition (e.g., poverty concentration), and school structure (e.g., size), organization, and climate (e.g., discipline) (see also Blue & Cook, 2004; Dalton et al., 2009; McNeal, 1997). Remarkably, a large proportion of relevant research remains focused on factors related to students, rather than on those related to schools (De Witte et al., 2013; Lee & Burkam, 2003). Therefore, it remains unclear whether and how the organizational and functional features of schools, for example, aspects of culture and climate are associated with dropout, although these are alterable and therefore potentially useful with regard to prevention (Luyten et al., 2003). Teachers in particular have been paid relatively little attention, notwithstanding evidence that the quality (as reported by students) and quantity (student-teacher ratios) of teachers have an effect on dropout ratios (Rumberger & Thomas, 2000). Moreover, a school's social climate, specifically student-teacher interactions and relationships as perceived and reported by students, has been demonstrated to be associated with dropping out. Students' reports of positive, caring, and supportive relationships with teachers coincide with lower dropout rates (Barile et al., 2012; Blue & Cook, 2004; Croninger & Lee, 2001; Lamote et al., 2013; Lee & Burkam, 2003). Accordingly, relationships perceived as negative, or student-teacher conflicts, might push students out of school, whereas positive relationshipsesocial capitalemight create
bert & powerful incentives to stay in school and be successful (He Reis, 1999; Lee & Burkam, 2003; Stearns & Glennie, 2006). 2.2. Teachers' beliefs and students' dropout Research into dropping out points to the fact that it is not only associated with problems regarding learning and academic engagement, but also with problems regarding social engagement (Finn, 1989; Rumberger & Palardy, 2005; Wehlage & Rutter, 1986). In this vein, Finn (1989) presented the ‘participation-identification model’, which emphasizes the importance of ‘bonding’ with school. If this bonding does not occur, the likelihood of problem behavior, including leaving school before graduation, increases. Studies into the impact of the social climate of schools, however, commonly rely on students' reports and perceptions of the studenteteacher relationships and the support from teachers (e.g., Barile et al., 2012; Fall & Roberts, 2012; Lamote et al., 2013; Lee & Burkam, 2003). This might be deceptive, particularly in cross-sectional studies, as the perceptions of students who drop out might be formed after they leave school. Furthermore, these perceptions might express negative feelings toward school irrespective of how teachers act in reality, and therefore might not provide insight into the actual role of teachers in the dropout process. A more useful and accurate indicator of the quality of teacherestudent relationships, not obtained from the students themselves, may be teachers' beliefs or expectations about their pupils (Van Houtte, 2011). After all, how teachers relate to and interact with their students is largely informed by how they see these pupils and what they think about them (Van Houtte, 2004, 2011). Over the years, colleagues (in the same school, for example) develop common ideas and

M. Van Houtte, J. Demanet / Teaching and Teacher Education 54 (2016) 117e127

views to react to the circumstances and problems that are particular to their work. In this regard, teachers and principals share certain beliefs about the nature of students, education, and school, forming teacher cultures (Hargreaves, 1992). The teacher culture consists of the work-related beliefs and knowledge teachers share; beliefs about appropriate ways of acting in the job, about the rewarding aspects of teaching, and about the knowledge that enables them to do their work, for example, ideas about who the students are and what they are capable of (Feiman-Nemser & Floden, 1986). Since the 1960s, educational researchers have been in agreement that the opinions teachers have about students can have a profound impact on students' educational progress (e.g., for the USA: Brophy & Good, 1970; Jussim & Harber, 2005; Rosenthal, 2002. For France: Trouilloud, Sarrazin, Martinek, & Guillet, 2002; for New Zealand: Rubie-Davies, Hattie, & Hamilton, 2006). Rosenthal and Jacobson (1968) were the first to present evidence regarding self-fulfilling prophecies in education. Specifically, they contended that students bring particular characteristics to the school context, which are, mostly unwittingly, used by teachers as an indication of the students' potential educational success. Rosenthal and Jacobson's (1968) main contention was that teachers' expectations determine their behavior toward students, which can actually result in raising students' performance. Pupils who teachers label as ‘gifted’ make the greatest progress, primarily because of differential treatment by teachers (Jussim, 1986; Rubovits & Maehr, 1971). It has been suggested that the attitudes of teachers shape their treatment of students in two ways (Jussim, 1986; Rosenthal, 2002). First, when their expectations of some students are low, they spend less effort and time teaching those pupils (Jussim, 1986). Second, lower expectations result in lesssupportive teacherestudent relationships (Jussim, 1986; Rubovits & Maehr, 1971). The Pygmalion study raised considerable controversy and resulted in much research into the effects of teacher expectations (e.g., Hinnant, O'Brien, & Ghazarian, 2009; Hughes, Gleason, & Zhang, 2005), most of which focused on students' cognitive outcomes (for reviews, see Brophy, 1983; Jussim & Harber, 2005). A few studies have also focused on non-cognitive outcomes, such as school misconduct (Demanet & Van Houtte, 2012) or attachment to school (Hallinan, 2008). Hallinan (2008) investigated the effect of teachers' expectations, as perceived by students, on the students' enjoyment of school. She hypothesized that students would enjoy school more if they could live up to the expectations of their teachers, but actually found that these perceived expectations did not affect students' liking of school. In a related study, Hinojosa (2008) looked at punitive actions against students related to outcomes of teacher expectations. She found that students who reported higher teacher expectations had a lower likelihood of being suspended from school, and speculated that students are more attached to school if they feel that teachers have high expectations of them. However, both the studies by Hallinan and Hinojosa were based on students' perceptions regarding the expectations of teachers. Rumberger and Palardy (2005) showed that schools where teachers had high expectations (measured among teachers themselves) had lower dropout rates. Moreover, Demanet and Van Houtte (2012) demonstrated that lower expectations gave rise to more oppositional behavior. As disorderly behavior at school is strongly related to lower achievement (Bryant, Schulenberg, Bachman, O'Malley, & Johnston, 2000) and dropout (Jenkins, 1995), there are certainly reasons to expect that low teacher expectations or poor teacher beliefs in students might result in worse academic progress and eventually students dropping out.

119

2.3. Vocational education 2.3.1. Dropping out in lower tracks In many countries, students in secondary education are grouped according to their level of ability, which is a practice founded on the belief that students have relatively fixed levels of ability and need to be taught accordingly (Boaler, William, & Brown, 2000; Gamoran, Nystrand, Berends, & LePore, 1995). This grouping of ability is organized in a myriad of ways. ‘Tracking’ refers to a situation in which students are taught an entirely different curriculum depending on their ability group, with the different tracks commonly classified hierarchically in terms of level of abstraction and theorizing, placing technical and vocational tracks at the bottom of the ladder (Boaler et al., 2000). Technical or vocational training is therefore usually not a positive choice, but instead a second choice because of not meeting the standards set for academic tracks (Ainsworth & Roscigno, 2005). This is certainly the case in educational systems where curriculum placement is based on prior achievement, such as in Flanders (Trautwein, Lüdtke, € ller, & Baumert, 2006). A lack of perceived ability, based Marsh, Ko on achievement, forces students into lower tracks, which entails a loss of status due to the hierarchical nature of the system (Hargreaves, 1967; Rosenbaum, 1976). Moreover, in the contemporary knowledge society, the occupations for which students are prepared in technical or vocational tracks are often held in low esteem, and de-industrialization and technological change have led to a collapse in the demand for skilled, semi-skilled, and unskilled manual workers (Nixon, 2006). The unemployment rate increases as the educational level decreases, offering technical or vocational students poor future prospects. Technical and vocational tracks seem to suffer from a negative image, resulting from the social overvaluing of cognition and white-collar jobs at the expense of manual labor. Given this undervaluation within society, it is conceivable that lower-track students lose faith in the system and no longer see the point of studying or working hard at school. Ethnographic case studies have pointed to the existence of more fatalism in lower tracks. For example, Schafer and Olexa (1971) suggested that lower-track students develop more negative attitudes toward school, partly because they consider that grades, commitment, and staying in school until graduation will show few future returns. Malmberg and Trempala (1997) indeed found that vocational-track students express lower levels of control over their future than academic-track students do. Those in the more academically-oriented tracks are, according to Friedkin and Thomas (1997), on average less fatalistic and more self-efficacious than students in vocationally-oriented tracks. This fatalism is likely to negatively affect the willingness of lower-track students to deliver effort at school (Carbonaro, 2005; Rosenbaum, 2001). Previous research in Flanders has shown that students in technical and vocational tracks display higher feelings of futility than students in academic tracks do (Van Houtte & Stevens, 2008, 2010, 2015), leading to less study involvement (Van Houtte & Stevens, 2010) and higher levels of school misconduct (Van Houtte & Stevens, 2008). It has repeatedly been shown that lower-track students tend to perform worse, achieve less, fail more frequently, and be more prone to dropping out compared with students in higher tracks (e.g., Duru-Bellat & Mingat, 1997; Hallinan & Kubitschek, 1999; Natriello, Pallas, & Alexander, 1989; Shavit & Featherman, 1988). A number of studies have suggested that students make the decision to drop out based on their perception of their future opportunities and the utility of education (Bickel, 1989; Worrell & Hale, 2001). Therefore, it is easy to understand why dropout is most prevalent in the lower tracks, specifically in vocational education (for Flanders: Van Landeghem & Van Damme, 2011; for the Netherlands: van Uden et al., 2014; for Turkey: Tas et al., 2013).

120

M. Van Houtte, J. Demanet / Teaching and Teacher Education 54 (2016) 117e127

Moreover, among vocational students, it might be expected that the most fatalistic and pessimistic have the highest likelihood of dropping out. Tas et al. (2013) found that most vocational students drop out because of social and academic factors, such as grade retention. However, most participants in their research also pointed to the attitudes and behavior of teachers, stating that teachers were unfair in their practices. 2.3.2. Teachers in the lower tracks It has been shown that teachers in lower tracks tend to share negative views about students; lower-track students are deemed less able, more disruptive, and less interested in schooling (Stevens & Vermeersch, 2010), and therefore less teachable and less trustworthy (Van Houtte, 2004, 2006a, 2006b). When teachers end up in tracks characterized by specific groups of students, they need to adjust their general conceptions about teaching to the real context, which might deviate from what they learned in teacher training (Fang, 1996). To do this, many will appeal to common stereotypes. Accordingly, the allocation of students into tracks is a given for these teachers, who usually do not make independent, individual evaluations of students, but start from the stereotype that lowertrack students are academically lacking (Ball, 1981; Rosenbaum, 1976). In other words, even before they have met their classes, many teachers have formed an image of their students' academic abilities and have developed certain expectations, in line with which they adjust their goals for, and interactions with, students (Ball, 1981; Finn, 1972; Jussim, 1986; McLaughlin, 1993; Metz, 1993; Midgley, Feldlaufer, & Eccles, 1988; Page, 1991). As a further result, teachers in low-track classes often demand less academically (Boaler et al., 2000; Evertson, 1982; Goodlad, 1984; Hargreaves, 1967; Oakes, 1985, 2005; Page, 1991; Stevens & Vermeersch, 2010). Generally speaking, the attitude of many teachers in higher tracks is more in tune with promoting learning than it is in lower tracks (Oakes, 1985, 2005; Van Houtte, 2004, 2006a), although there are also schools with apparently effective instruction in low tracks, characterized by high expectations from teachers (see Gamoran, 1993). Kelly and Carbonaro (2012) demonstrated in a system of withinschool tracking that track placements affect teacher expectations above and beyond student achievement and other characteristics. However, in a system of between-school tracking, Van Houtte, Demanet, and Stevens (2013) showed that teachers' negative judgments of students in lower tracksemore specifically in technical and vocational schools compared with academic schoolsecould be ascribed to the students' background variables, namely gender, socioeconomic status, immigrant status, and especially, ability. Teachers' views of students are more negative in lower tracks because teachers base their judgment on precisely the characteristics that make students opt for the lower tracks in the first place. However, irrespective of their origin, teachers' shared beliefs and expectations have been demonstrated to be responsibleeat least in partenot only for the higher failure rate (Van Houtte, 2004), but also for the lower sense of school belonging (Van Houtte & Van Maele, 2012) of technical and vocational students. Knowing that many teachers in lower tracks hold a relatively poor image of their students, and given that lower track students are most vulnerable to dropping out, an obvious question is to what extent, especially in vocational tracks, is the intention of students to drop out associated with teachers' shared beliefs and expectations?

subsidized; private as well as state schools. Usually, children attend nursery school from the age of two and a half. Education becomes compulsory when the child turns six years of age. After six years of primary education, children transfer to secondary education at the age of twelve. There are then six years of secondary education divided into three grades, each lasting two years. In theory, the first grade (years one and two) is an orientating grade officially divided into a core curriculum known as the A-stream, and a B-stream preparing for vocational education. In practice, however, the type of courses offered in the A-stream depend on the main tracks offered in the particular school. There are four main tracks: academic education preparing for higher education, technical education, vocational education, and artistic education (which is relatively marginal in terms of the number of students). In Flemish secondary education, tracks are organized within and, mainly, between schools. A common differentiation is between schools offering academic education and those offering technical and vocational education, but there are also schools offering just one track and others offering all the tracks; termed multilateral schools. Hence, vocational students might be enrolled in a school offering solely vocational education, or in a school also offering academic tracks. The latter has been shown to be the most influential for vocational students, since in these multilateral schools vocational students compare themselves with academic track students more directly, resulting in stronger antischool attitudes, namely lower study involvement and higher sense of futility (Van Houtte & Stevens, 2009, 2015). Within each main track, different tracks are distinguished, for example, economy and modern languages in academic education, electricity and mechanics in technical education, and childcare in vocational educationecharacterized by different subjects and emphasis. At the end of each year, students are given a certificate indicating whether they can continue their current school path (certificate A) or not (certificate B or C). In the case of the latter two, a certificate B indicates that the student may progress to the next year but needs to join a lower track, and a certificate C means that the student cannot move on and has to repeat the year. These certificates are based on the obtained Grade Point Average (GPA), which is determined by the teachers based on tests and assignments designed by the teachers. There are no standardized tests, for example in the form of centrally-administered and standardized examinations (Stevens, 2007). At the end of each grade, that is, in the third and the fifth year, the students need to refine their branch of studies. There is an option to enroll in parttime vocational education from the age of sixteen, combining classes with experience on the shop floor. After six years of general, technical, or artistic education, or seven years (six years plus an extra year) of vocational education, the student receives a diploma of secondary education, granting unlimited access to each form of higher education. Each student having a diploma of secondary education may enter university. If they only complete six years of vocational education (without the additional year), students obtain a qualification that does not grant access to higher education. Secondary education is compulsory until the age of eighteen, and a student is considered to have dropped out if they leave education before having finished the six years and having obtained a diploma (academic, arts, and technical education) or qualification (vocational education). 4. Methods

3. Context

4.1. Data

Before describing the methodology of the study, it seems useful to describe briefly the Flemish educational system. First of all, it should be kept in mind that every school in Flanders is state

Our analyses are based on data from the Flemish Educational Assessment (FlEA), gathered in 2004 and 2005 from 11,872 third and fifth grade students (equivalent to ninth and eleventh grade

M. Van Houtte, J. Demanet / Teaching and Teacher Education 54 (2016) 117e127

students in the American education system) clustered in a representative sample of 85 secondary schools in Flanders. Schools agreeing to participate did so with parental consent. Students completed the questionnaires in class, in the presence of one or two researchers and a teacher. In the end, 11,945 students completed a questionnaire, of which 11,872 proved to be valid; a response rate of 87%. As part of the FlEA, a questionnaire was also distributed among all third and fifth grade teachers in the selected schools. A total of 2104 teachers (response rate ¼ 60%) returned the questionnaire. However, in one school, no teachers responded to the questionnaire. As we used aggregated data to estimate the effects of teacher beliefs, and multilevel analysis does not permit missing values at the school level, we had to remove this school and its students from the analyses. Hence, the analyses were performed for a total of 11,844 students across 84 schools. The questionnaires were not anonymous, because we wanted to couple the data with other information such as academic results provided by the schools. All the names were removed as the data were assembled, so the final dataset and all the analyses are completely anonymous. The analyses for the vocational tracks were based on 2589 students in 48 schools offering vocational education (for the descriptives, see Table 1). 4.2. Design As the data is cross-sectional, gathered within schools, there is no information on the actual number of students who dropped out. However, we did ask the students whether they were planning to complete high school. If not, they were assumed to have been intending to quit, although we are not aware of whether these students actually did so. It is most plausible that some of them in the end did stay on in education. Expressing an intention to quit is nonetheless an indication of severe disengagement (Janosz, Archambault, Morizot, & Pagani, 2008). Given the clustered sample of students nested within schools, and with an outcome at the student level (the intention to drop out) and determinant at the school level (teachers' shared teachability beliefs, that is, a schools' teachability culture), multilevel techniques were most appropriate. In view of the binary outcome (intending to drop out ¼ 1, not intending to ¼ 0), we used nonlinear Bernoulli models (HLM6, Bryk & Raudenbush, 1992). Multilevel analyses commonly start by estimating unconditional models to

121

determine the amount of variance that occurs; in this case, between schools. However, in hierarchical logistic models, it is not appropriate to partition the variance of the outcome into its between and within components. The between-school variance component t0 estimated in an unconditional model does give an idea of whether or not the between-school variance is significant and can be modeled (e.g., Frost, 2007). In a first model, we included individual characteristics that have been shown to be related to dropout: social risk factors (gender, age, socioeconomic status, and immigrant background) and academic risk factors (ability, achievement, being retained, and track position). In the second model, we added teachability culture (the shared expectations of teachers regarding students' teachability) at the school level, to assess the role of teachers in the intentions of students to drop out. In the third model, we added students' perception of teacher support to assess whether this explained an eventual association between the teachability culture and the intention to dropout. This was a possibility, in light of existing research indicating that lower teacher expectations lead to lower perceived support. In the fourth model, we added the culture of futility (school level) and the sense of futility (student level), to ascertain whether an association between teachability and the intention to drop out still remained. This would indicate that the expectations of teachers were not merely a response to students' negative attitudes, but that teachers' expectations actually affected the intention of students to drop out. These four models were repeated for vocational students, although omitting the track position variables in the first model and taking into account (at the school level) whether or not the vocational track was in a school that also offered academic tracks (school type), as this might have affected the teachability culture. After all, teachers were asked how they perceived the students at their school in general (see Section 4.3.), which might differ between schools offering solely vocational education and those offering vocational and academic education. 4.3. Variables 4.3.1. Outcome As our data concerns students in a sample of schools, we could not take into account students who had already dropped out. Instead, we asked the students whether or not they were planning

Table 1 Intention to dropout in secondary education in Flanders: Descriptives. Total sample

School level Teachability culture Futility culture School type: academic Student level Dropout intention Gender: girls Age SES Immigrant background Ability Achievement Retained Academic track Arts track Technical track Vocational track Teacher support Sense of futility

Vocational students

Frequency (%) or mean (SD)

N

Frequency (%) or mean (SD)

N

99.79 (10.35) 10.06 (0.73)

84 84

93.46 (7.89) 10.46 (0.66) 22.9%

48 48 48

5.4% 51.5% 16.45 (1.31) 5.21 (2.10) 11.1% 77.98 (9.85) 69.42 (9.22) 20.3% 46.8% 2.7% 28.5% 21.9% 23.99 (3.98) 9.99 (3.20)

11,519 11,815 11,775 11,111 11,842 10,673 10,685 11,517 11,844 11,844 11,844 11,844 11,593 11,587

13.9% 53.4% 16.95 (1.30) 3.58 (2.02) 25.3% 70.12 (10.62) 69.29 (10.40) 27.2%

2437 2589 2579 2210 2598 2008 2105 2433

23.54 (4.29) 10.88 (3.55)

2467 2466

122

M. Van Houtte, J. Demanet / Teaching and Teacher Education 54 (2016) 117e127

to complete high school with a credential. The majority of the students (94.6%) were aspiring to finish high school (coded 0), 5.4% of the total sample and 13.9% of the vocational students expressed the intention to quit (see Table 1). 4.3.2. School-level variables To investigate the beliefs of teachers about the pupils at their school, we used the ‘Teachable Pupil Survey’ developed by Kornblau (1982) (see also Demanet & Van Houtte, 2012). This fivepoint scale comprises 31 items, encompassing ‘school-adjusted behaviors’ (e.g., ‘concentrates well’ and ‘enjoys school work’), ‘cognitive-motivational behaviors’ (e.g., ‘intelligent’ and ‘curious, inquisitive’), and ‘personalesocial behaviors’ (e.g., ‘calm’ and ‘confident’). The score for each teacher's answers was computed by summing the items, with missing ones being imputed by item correlation substitution (Huisman, 2000). This yielded a range from 39 to 146, with a mean score of 100.15 (SD ¼ 15.307; Cronbach's alpha ¼ 0.94; n ¼ 2,104). By summing the item scores, we obtained a reliable measurement of perceived teachability at the individual teacher level. However, because the purpose was to measure culture, and because culture is a group feature defined as ‘shared beliefs,’ aggregation of the scores obtained was a necessary subsequent step. One customary aggregation strategy is to calculate the mean of the scores of the individual members of a group or organization (e.g., Hofstede, Neuijen, Ohavy, & Sanders, 1990). In this process, it is necessary to ensure this aggregation is valid by ascertaining whether the aggregated measurement is reliable and represents something actually shared at the group or organization level (Glick, 1985). One useful measurement is the ‘mean rater reliability’ (Glick, 1985), calculated by means of the SpearmanBrown formula based on the intraclass correlation (ICC) of a oneway analysis of variance: ICC(1, k) ¼ (between mean squarewithin mean square)/between mean square (with k ¼ number of raters in each group or organization). The result must reach a minimum of 0.60 to permit aggregation at the group or organization level (Glick, 1985; Shrout & Fleiss, 1979). Aggregation of this measurement at the school level, by calculating the mean per school, was legitimate (ICC ¼ 0.92). For the 84 schools in the sample, the measurement for the culture of teachability had a mean of 99.79 (SD ¼ 10.35, Table 1), for the 48 schools attended by vocational students this was 93.46 (SD ¼ 7.89). To assess the schools' culture of futility a similar procedure was followed, starting with the measurement of students' feeling of futility (see below). Aggregation of this measurement at the school level by calculating the mean per school was valid (ICC ¼ 0.84). For the 84 schools in the sample, the score for the culture of futility had a mean of 9.99 (SD ¼ 3.20, Table 1), for the 48 schools attended by vocational students, this was 10.88 (SD ¼ 3.55). In the analyses for vocational students, we added (at the school level) whether the school offered an academic track (coded 1, see Section 4.2.). Of the 48 schools attended by vocational students, 22.9% offered an academic track. 4.3.3. Student-level variables The sample was almost equally divided with regard to gender, with 51.5% of the total sample and 53.4% of the vocational students being female (coded 1). Our research concentrated on third and fifth-grade students, so the respondents were on average 16.45 years old (SD ¼ 1.31), although the average age of the vocational students was slightly above this (mean ¼ 16.95, SD ¼ 1.30), due to a higher retention rate (Table 1). We measured the socioeconomic status (SES) of origin of the students by means of the occupational prestige of the father and mother (Erikson, Goldthorpe, & Portocarero, 1979), with the highest of the two used as an indicator of the family SES. The

respondents showed a mean SES of 5.21 (SD ¼ 2.10, range 1e8). On average, the vocational students had a lower SES (mean ¼ 3.58, SD ¼ 2.02, Table 1). We also distinguished between native and immigrant students. As is common (Timmerman, Hermans, & Hoornaert, 2002), the principal criterion was the birthplace of the students' maternal grandmother. If this was missing (only 1% missing of the total sample, n ¼ 11,872) we used the nationality of the mother and father, as most immigrant students are second or third-generation and have Belgian nationality. As is normal practice in Flemish and Dutch educational research (Duquet, Glorieux, Laurijssen, & Van Dorsselaer, 2006; Sierens, Van Houtte, Pelleriaux, Loobuyck, & Delrue, 2006; Timmerman et al., 2002), birthplaces and nationalities other than Western European are considered as foreign descent, because the relevant students are more likely to face educational difficulties. Additional criteriaein the event of missing data regarding nationalityewere the language spoken at home (a language other than Dutch), religion, and finally the student's name (see Felouzis, 2003). We created a dichotomous variable (0 ¼ indigenous, 1 ¼ immigrant). In the available data, 11.1% of the students were identified as being of foreign origin. The proportion of immigrant students who were in the vocational tracks was higher (25.3%, Table 1). As a measurement of ability, we used the GPA at the completion of primary education (range 0e100). This requires caution. There are no standardized tests in the form of centrally-administered and standardized examinations in the Flemish educational system, which makes educational achievement very hard to compare across schools and across students. Teachers are responsible for designing tests and grading students, generating the GPA (Stevens, 2007). Furthermore, we need to rely on a self-reported GPA, resulting in issues concerning validity due to memory problems and cover-up strategies. Not unexpectedly, this variable showed more missing values than other variables do (9.2%). Notwithstanding these problems, it remained the best measurement available to us to control for prior academic attainment. The mean GPA for the total sample was 77.98 (SD ¼ 9.85). Vocational students showed a lower average GPA (mean ¼ 70.12, SD ¼ 10.62, Table 1). To measure prior academic achievement, we used the GPA at the end of the grade preceding the survey (range 0e100, pass mark of 50). Again, due to the lack of centrally-administered standardized tests, we also needed to rely on self-reported GPA for this variable, resulting in the same issue of validity as applies to ability. However, research indicates that self-reported grades are generally highly correlated with grades taken from students' transcripts, and that GPA has some desirable features relative to standardized test scores (Kelly, 2008). The mean GPA for the total sample was 69.42 (SD ¼ 9.22). The average for vocational students was lower (69.29, SD ¼ 10.40, Table 1). We measured retention by asking the respondents to report retrospectively on their history of grade retention (0 ¼ never retained; 1 ¼ retained at least once). Of the respondents in the total sample, 20.3% indicated that they had been retained at least once in the course of primary and secondary education. For the vocational students, the corresponding figure was 27.2%. Four tracks can be distinguished in the Flemish educational system: the academic track (46.8% of the students in our sample), the arts track (2.7%), the technical track (28.5%), and the vocational track (21.9%). To account for the students' track position, the last three were added to the analyses by means of dummy coding, with academic track as the reference category. Perceived teacher support was measured by a subscale of the Psychological Sense of School Membership (PSSM) scale (Goodenow, 1993). The scale consists of seven items, such as ‘Teachers in this school respect me,’ and ‘Teachers in this school are

M. Van Houtte, J. Demanet / Teaching and Teacher Education 54 (2016) 117e127

not interested in students like me.’ Students could choose from five answers, ranging from ‘completely disagree’ to ‘totally agree’ (1e5). Scores across the items were summed, yielding a scale ranging from 7 to 35 (Cronbach's alpha ¼ 0.74; n ¼ 11,593). For the total sample, the mean was 23.99 (SD ¼ 3.98; see Table 1), for vocational students it was 23.54 (SD ¼ 4.29). Sense of futility was measured by means of a scale adapted from Brookover, Beady, Flood, Schweitzer, and Wisenbaker (1979), and consisting of five items dealing with the feeling of control at school, such as ‘People like me will not have much chance to do what we want to in life.’ Each item had five answer categories, ranging from ‘completely disagree’ to ‘totally agree’ (range 1e5). Responses were imputed for missing values by way of item correlation substitution (Huisman, 2000). In this analysis we worked with the sum of the item scores, yielding scores from 5 to 25 (Cronbach's alpha ¼ 0.75, n ¼ 11,620). In the total sample, the mean was 9.99 (SD ¼ 3.20), while the mean for vocational students was higher (10.88, SD ¼ 3.55, Table 1). 5. Findings The between-school variance in the unconditional model indicated that it was useful to estimate a model taking into account school-level variables (t0 ¼ 0.774, p < .001). The first model confirmed the impact of the social and academic risk factors, except for migrant status, which proved not to be significantly related to the intention to quit (Table 2). With regard to the social risk factors, boys, younger students, and students with a lower socioeconomic status were significantly more likely to indicate that they considered dropping out, although the differences were small. For the academic risk factors, students with a lower ability and with a lower achievement showed a significantly higher intention to quit. The most important predictors appeared to be whether the students had been retained previously, and the track. The vulnerable position of vocational students was confirmed: if academic students had a probability of 2.25% of intending to quit, the equivalent figure was 8.96% for vocational students (3.59% for technical students). In the second model, teachability culture was borderline (p < .10, see Lee & Burkam, 2003: 373) significantly and negatively (OR ¼ 0.984, p ¼ .07) associated with the intention to quit (Model 2,

123

Table 2). This small association decreased slightly (OR ¼ 0.986) and was no longer borderline significant (p ¼ .134) when taking into account students' perception of teacher support (OR ¼ 0.880, p < .001) (Model 3, Table 2), meaning that the effect of teachers' culture was mediated by students' perceptions of teacher support. When adding culture and sense of futility in the final model, the association between teachability culture and intention to quit increased slightly (OR ¼ 0.981) and again became borderline significant (p ¼ .089), although remaining very small. In the final model, the significant predictors of the intention to drop out turned out to be the school's teachability culture and student's perceived teacher support, in addition to the expected gender, socioeconomic status, achievement, retention, vocational track position, and sense of futility. It is noteworthy, therefore, that the difference between students in academic and vocational tracks was not due to a difference in teacher culture. For the vocational track, the between-school variance in the unconditional model indicated that it was useful to estimate a model taking into account school-level variables (t0 ¼ 0.059, p ¼ .006). Surprisingly, none of the social or academic risk factors appeared significantly related to the intention to drop out, indicating that once enrolled in the vocational track, these characteristics apparently no longer mattered or made a difference. One exception was achievement, which was borderline significant (OR ¼ 0.984, p ¼ .057; Table 3, model 1). Teachability culture proved significantly negatively related to the intention to quit (Model 2: OR ¼ 0.974, p < .05) and this association held after controlling for perceived teacher support, teachability culture, and sense of futility (Models 3 and 4). The most important predictor of intending to drop out for vocational students proved to be their sense of futility (OR ¼ 1.173, p < .001), followed by perceived teacher support (OR ¼ 0.951, p < .01), and the teachability culture of the school (OR ¼ 0.972, p < .05). 6. Discussion The educational level of a country is one of the most important indicators of its socioeconomic development and productivity rates. A high dropout rate entails a loss of human capital, and affects policymakers' objectives for sustainable economic growth (De

Table 2 Correlates of intention to dropout in secondary education. Results of logistic multilevel analysis, Bernoulli (HLM 6).

Intercept School level Teachability culture Futility culture Student level Gender Age SES Immigrant background Ability Achievement Retained Arts track (ref. Academic) Technical track (ref. Academic) Vocational track (ref. Academic) Teacher support Sense of futility Variance components Intercept U0 Teacher support U1

Model 1

Model 2

Model 3

Model 4

3.774 (0.160) 0.023***

3.641 (0.178) 0.026***

3.811 (0.168) 0.022***

3.894 (0.172) 0.020***

0.017 (0.009) 0.984

0.014 (0.009) 0.986

0.019 (0.011) 0.981þ 0.166 (0.138) 0.848

0.335 (0.104) 0.715** 0.104 (0.047) 0.901* 0.123 (0.027) 0.885*** 0.012 (0.167) 1.012 0.015 (0.005) 0.986** 0.024 (0.006) 0.976*** 0.451 (0.149) 1.571** 0.261 (0.269) 1.298 0.484 (0.189) 1.622* 1.456 (0.197) 4.289***

0.311 (0.104) 0.732** 0.106 (0.047) 0.900* 0.118 (0.026) 0.889*** 0.051 (0.170) 0.951 0.014 (0.005) 0.986* 0.025 (0.006) 0.975*** 0.433 (0.150) 1.542** 0.127 (0.277) 1.135 0.301 (0.213) 1.352 1.239 (0.243) 3.453***

0.256 (0.110) 0.774* 0.110 (0.049) 0.896* 0.112 (0.027) 0.894*** 0.023 (0.180) 0.978 0.013 (0.005) 0.987* 0.021 (0.006) 0.980** 0.439 (0.144) 1.552** 0.023 (0.227) 1.024 0.307 (0.200) 1.024 1.250 (0.236) 3.490*** 0.128 (0.014) 0.880***

0.262 (0.113) 0.769* 0.074 (0.050) 0.929 0.104 (0.028) 0.901*** 0.064 (0.205) 0.938 0.011 (0.006) 0.989 0.014 (0.006) 0.987* 0.387 (0.147) 1.472** 0.046 (0.244) 0.955 0.294 (0.205) 1.341 1.168 (0.245) 3.216*** 0.081 (0.016) 0.923*** 0.180 (0.017) 1.197***

0.042*

0.036

0.041 0.003*

0.044 0.005*



Note: Presented are the (unstandardized) gamma coefficients with the standard errors appearing in parentheses and odds ratios and the variance components U (when significant). þ p ¼ .089,  p < .07, *p < .05, **p  .01,***p  .001.

124

M. Van Houtte, J. Demanet / Teaching and Teacher Education 54 (2016) 117e127

Table 3 Correlates of intention to dropout in vocational secondary education. Results of logistic multilevel analysis, Bernoulli (HLM6).

Intercept School level School type Teachability culture Futility culture Student level Gender Age SES Immigrant background Ability Achievement Retained Teacher support Sense of futility Variance components Intercept U0

Model 1

Model 2

Model 3

Model 4

2.163 (0.134) 0.115***

2.193 (0.127) 0.112***

2.278 (0.138) 0.103***

2.308 (0.136) 0.100***

0.208 (0.251) 1.232 0.026 (0.012) 0.974*

0.159 (0.269) 1.173 0.025 (0.012) 0.976þ

0.048 (0.264) 1.050 0.028 (0.014) 0.972* 0.148 (0.171) 0.862

0.084 (0.164) 0.920 0.045 (0.052) 0.957 0.077 (0.047) 0.926 0.289 (0.213) 1.335 0.010 (0.008) 0.990 0.016 (0.008) 0.984 0.245 (0.229) 1.278

0.033 (0.174) 0.968 0.035 (0.052) 0.966 0.067 (0.047) 0.935 0.174 (0.212) 1.190 0.009 (0.008) 0.991 0.016 (0.008) 0.984* 0.214 (0.228) 1.239

0.048 (0.185) 1.049 0.032 (0.052) 0.968 0.055 (0.047) 0.947 0.200 (0.217) 1.222 0.009 (0.008) 0.992 0.015 (0.008) 0.985 0.224 (0.222) 1.251 0.092 (0.016) 0.913***

0.053 (0.179) 1.054 0.002 (0.054) 1.002 0.053 (0.050) 0.949 0.149 (0.244) 1.160 0.006 (0.008) 0.994 0.011 (0.008) 0.989 0.154 (0.220) 1.166 0.050 (0.017) 0.951** 0.159 (0.023) 1.173***

0.008

0.034

0.052

0.077

Note: The (unstandardized) gamma coefficients and odds ratios with the standard errors appearing in parentheses and the variance components U (when significant) are shown.  p ¼ .057, þp ¼ .052, *p < .05, **p  .01, ***p  .001.

Witte et al., 2013; Lee & Burkam, 2003; Tas et al., 2013). On a more personal level, dropping out is a major life event that severely impacts students' chances for ensuing educational and occupational opportunities (Dalton et al., 2009; Rumberger, 1987). Because students dropping out without qualifications is manifestly an important problem within Western education (Lamb et al., 2011), a great deal of research has been carried out, aiming to reveal the main predictors of student dropout. However, most research has focused on individual student characteristics, specifically sociodemographic and academic risk factors (De Witte et al., 2013), turning dropout into an individual problem (Luyten et al., 2003). In particular, research into the role of teachers in students' decision to quit school is relatively scarce. Research taking teachers into consideration often focuses on students' perceptions of teacher support or studenteteacher relationships (Croninger & Lee, 2001; Lee & Burkam, 2003), neglecting that these perceptions may be colored and may not accurately reflect the teachers' actual role. Therefore, this study had the objective of ascertaining whether teachers' shared beliefs regarding students were associated with the intention of students to quit school, irrespective of students' perceptions of teacher support. By means of multilevel analyses, the study showed that all the social and academic risk factors at the individual student level, and already defined in international research, are indeed significantly associated with students' intention to drop out. Male students of a younger age, with a low socioeconomic status, lower ability and lower achievement, and who had been retained at least once during their schooling, were more likely not to aspire to finish high school with any qualifications. The most important predictor, however, appeared to be track position; specifically, being enrolled in a vocational track. In addition, students' perceived teacher support and sense of futility were also significant predictors. The teachability culture in school, however, proved to be only borderline significantly related to students' intention to drop out, and this association acted through students' perceptions of teacher support. Hence, it appeared that associations between teacher attitudes and students' intention to drop out were smaller when teachers' selfreported attitudes were used than when using perceptions reported by students. The findings were in line with the vast body of literature on the beneficial impact of students' perceptions of having supportive relationships with their teachers (e.g., Barile et al., 2012; Croninger & Lee, 2001; Lamote et al., 2013), but they

also showed that once students' perceptions were controlled for, the actual teacher-reported beliefs still had a small impact. Studies maintain that expectancy effects occur to a lesser extent when students do not perceive differential treatment by teachers based on their expectations (Brattesani, Weinstein, & Marshall, 1984; Jussim & Harber, 2005). In line with literature on expectancy, we found that expectancy effects on dropout intentions were mediated by students' perceptions of teacher support: in other words, the affective component of differential treatment by teachers (see also Rosenthal, 2002). For decades, research into expectancy effects has urged teacher education programs to raise awareness among trainee teachers of the beliefs they bring into the classroom, as teachers are not always aware of the images they have of students and the interaction patterns they maintain with students following from those images (see also Demanet & Van Houtte, 2012; Jones, 1989). However, we may expect that, despite the good intentions of teacher education programs, a newly trained teacher entering a school may be socialized into the prevailing teacher culture in that school. The present study, for example, demonstrated the existence and the important role of the teachability culture at school. The socialization into such a teacher culture most likely takes place in informal situations, for example, in the teachers' lounge, where teachers share their experiences from the classroom and talk about their students (Ben-Peretz & Schonmann, 2000; Hargreaves, 1992). For individual teachers, it may prove very difficult to turn the tide in a school where school-wide expectations for the students are low. As such, teacher education programs need to make trainee teachers aware as well of the existence of specific teacher cultures in schools, usually related with the specific student composition of schools. The question remains, then, whether individual teachers are able to counteract the influence of the teacher culture at school. In that regard, it is noteworthy that the current study points to the perceived teacher support of the students as an important mediating factor. This could imply that when individual teachers provide students with a sense of support, this may help to reduce the impact of the wider teacher culture (see also Demanet & Van Houtte, 2012). As such, it is important that teachers know that they should develop supportive relationships with every single student, even though they may have low expectations of them. As vocational students appear the most vulnerable to planning not to finish secondary education, it is important to know which

M. Van Houtte, J. Demanet / Teaching and Teacher Education 54 (2016) 117e127

risk factors are at play in this specific group. An analysis for this group separately, demonstrated that the traditional social and academic risk factors no longer played a role once a student is enrolled in a vocational track. In vocational tracks, the likelihood of planning to quit was not dependent on gender, age, SES, immigrant status, ability, or previous retention. When taking into account students' sense of futility, even prior achievement was no longer a predictor for the intention of vocational students to drop out. This finding contradicts statements by, for example, Tas and colleagues (2013) that the main reasons for dropout in vocational high school were individual academic factors such as grade repetition and achievement. These findings may be due to the specific nature of the track system in Flanders. As already noted, tracks are hierarchically ordered, with the vocational tracks at the lowest level. The finding that academic risk factors had no effect within the vocational track, suggested that attending this track in Flanders is in itself an academic risk factor for considering dropping out. The students concerned, felt that they had failed in the educational system, and therefore they were more likely to have the intention of dropping out prematurely. In light of this, it was unsurprising that the most important predictor of vocational students' intention to drop out turned out to be their sense of futility. Students who had a stronger feeling that studying was of no use to people like them, were more likely to indicate that they had the intention of quitting. Moreover, students' perceived teacher support and the teachability culture were also related to the intention to quit, with the teachability culture retaining its impact even when controlling for students' perceptions of teacher support. In contrast to the general sample, it appeared that teachers' self-reported beliefs were important for student dropout in vocational education. This was a substantive finding, as it suggested that teachers had the power to counteract, albeit only to a certain extent, the influence of vocational track position. More specifically, when students in the vocational track were confronted with beneficial teacher expectations, they were less likely to consider dropping out early. Teachers, therefore, held the key in demonstrating to students that attending the vocational track was not equal to failing in the educational system. Teachers, however, could not fully eradicate the negative influence of vocational track enrollment. Although the teachability culture and students' perceptions of teacher support could make a difference for vocational students, it needs to be stressed that these did not explain why vocational students had a significantly higher likelihood of planning to drop out. The association between track position and the intention to quit remained, when taking into account the teachability culture and perceptions of support. Even the sense of futility could not explain why vocational students were more likely not to aspire to finish high school. The detrimental consequences of being in a low track prompt the question of whether hierarchically-ordered tracks should be abandoned, particularly as mixed-ability groups have been shown to be beneficial for low-ability students (Hallam & Ireson, 2006; Ireson & Hallam, 1999; Linchevski & Kutscher, 1998). Nevertheless, one reason for tracking is preparing students for different futures, since societies are as much in need of manual workers as of brainworkers, so it is not advisable to abandon tracking altogether. Postponing the educational choice is recommended, though, to turn the choice for technical and vocational tracks into a more positive and a conscious one. Currently Flemish students need to choose already at the age of 12, making this transition largely a parental decision. Technical and vocational education is usually only considered an option when a student does not meet the achievement standards for the academic track, making it a negative choice. Thus the choice for vocational education is forced upon the students. Moreover, research has shown that next to achievement the social background of students plays a determining role in

125

educational choice (Boone & Van Houtte, 2013). Accordingly, there is a profound need for a social (re)appreciation of technical and vocational tracks and occupations. Whilst there is a relatively large group of unskilled unemployed people who dropped out of technical or vocational tracks, in Flanders there is a shortage of skilled manual workers, such as gardeners, welders, butchers, carpenters, bricklayers. To solve this discrepancy, the negative image of technical and vocational tracks needs to be reversed. It is important to acknowledge some limitations of the current study. First, we should stress again that this was a cross-sectional study, which precludes the possibility of making causal claims. It is possible, therefore, that the direction of the effect was reversed; that the intention to drop out led to lower teacher support and a worse teacher culture. As discussed above, it may be the case that students having decided to drop out of education eventually report more negatively about school-related matters in general, and about their relationships with teachers in particular. We took this into account by investigating teacher-reported expectations, but we considered the role of student-reported teacher support as well, which admittedly proved the stronger predictor of the two. Moreover, we controlled for students' sense of futility, which enabled us to investigate whether teacher expectations were not merely a response to students' negative attitudes and instead that teachers' expectations really affected students' intentions to drop out. We suggest, however, that future longitudinal studies should try to replicate the findings in order to support this interpretation of the direction of the effects. Further, longitudinal research would allow investigation of actual dropout behavior, instead of the intention to quit. A second limitation of this study is the way in which we operationalized teacher expectations. By considering school-wide attitudes, we could not go into detail regarding the effects of individual teachers. A further objection concerning the operationalization of teacher effects was that we could not distinguish between individual students or different groups of students within the same school: the culture of teachability is a collective idea about the teachability of students as a group, and therefore pertains to all the students at a school. It would be beneficial to have data from each teacher about his or her expectations of each student. However, this method of data collection would be very extensive and demanding of teachers, and may not be feasible in secondary education (see also Van Houtte, 2011, p. 84). Moreover, we contend that operationalizing teachers' expectations by means of the teacher culture at school has its merits. In secondary schools, it is more logical to investigate the role of teacher culture than that of individual teachers' expectations, because students are taught by a number of different teachers during each school year. 7. Conclusion This study contributes to existing knowledge about the predictors of high school dropout by demonstrating that teachers' beliefs about the teachability of students influence students' intention to quit high school, irrespective of students' sense of futility and perceived teacher support. The teachability culture makes a difference in particular for the most vulnerable group; vocational students. This finding underlines the importance of the school, and more particularly the role of teachers, in explaining dropout in secondary education. References Ainsworth, J. W., & Roscigno, V. J. (2005). Stratification, school-work linkages and vocational education. Social Forces, 84, 257e284. http://dx.doi.org/10.1353/ sof.2005.0087.

126

M. Van Houtte, J. Demanet / Teaching and Teacher Education 54 (2016) 117e127

Baker, J. A., Grant, S., & Morlock, L. (2008). The teacher-student relationship as a developmental context for children with internalizing or externalizing behavior problems. School Psychology Quarterly, 23, 3e15. Ball, S. J. (1981). Beachside comprehensive: A case-study of secondary schooling. Cambridge, England: Cambridge University Press. Barile, J. P., Donohue, D. K., Anthony, E. R., Baker, A. M., Weaver, S. R., & Henrich, C. C. (2012). Teacher-student relationship climate and school outcomes: implications for educational policy initiatives. Journal of Youth and Adolescence, 41, 256e267. http://dx.doi.org/10.1007/s10964-011-9652-8. Ben-Peretz, M., & Schonmann, S. (2000). Behind closed doors. Teachers and the role of the teachers' lounge. Albany, NY: State University of New York Press. Bickel, R. (1989). Opportunity and high school completion. The Urban Review, 21, 251e261. Blue, D., & Cook, J. E. (2004). High school dropouts: can we reverse the stagnation in school graduation? Study of High School Restructuring, 1, 1e11. Boaler, J., William, D., & Brown, M. (2000). Students' experiences of ability grouping: disaffection, polarisation and the construction of failure. British Educational Research Journal, 26, 631e646. http://dx.doi.org/10.1080/713651583. Boone, S., & Van Houtte, M. (2013). In search of the mechanisms conducive to class differentials in educational choice: a mixed method research. Sociological Review, 61, 549e572. http://dx.doi.org/10.1111/1467-954X.12031. Brattesani, K. A., Weinstein, R. S., & Marshall, H. H. (1984). Student perceptions of differential teacher treatment as moderators of teacher expectation effects. Journal of Educational Psychology, 76, 236e247. http://dx.doi.org/10.1037//00220663.76.2.236. Brookover, W. B., Beady, C. H., Flood, P. K., Schweitzer, J. H., & Wisenbaker, J. M. (1979). School social systems and student achievement. Schools can make a difference. New York: Praeger Publishers. Brophy, J. E. (1983). Research on the self-fulfilling prophecy and teacher expectations. Journal of Educational Psychology, 75, 631e661. http://dx.doi.org/10.1037/ 0022-0663.75.5.631. Brophy, J. E., & Good, T. L. (1970). Teachers communication of differential expectations for children's classroom performance: some behavioral data. Journal of Educational Psychology, 61, 365e374. http://dx.doi.org/10.1037/h0029908. Bryant, A. L., Schulenberg, J., Bachman, J. G., O'Malley, P. M., & Johnston, L. D. (2000). Understanding the links among school misbehavior, academic achievement, and cigarette use: a national panel study of adolescents. Prevention Science, 1, 71e87. Bryk, A. S., & Raudenbush, S. W. (1992). Hierarchical linear models. Applications and data analysis methods. Newbury Park: Sage Publications. Carbonaro, W. J. (2005). Tracking, students' effort, and academic achievement. Sociology of Education, 78, 27e49. Croninger, R. G., & Lee, V. E. (2001). Social capital and dropping out of high school: benefits to at-risk students of teachers' support and guidance. Teachers College Record, 103, 548e581. http://dx.doi.org/10.1111/0161-4681.00127. Dalton, B., Glennie, E., Ingels, S. J., & Wirt, J. (2009). Late high school dropouts: Characteristics, experiences, and changes across cohorts (NCES 2009-307). Washington, DC: National Center for Education Statistics, Institute of Education Sciences, U.S. Department of Education. Demanet, Jannick, & Van Houtte, Mieke (2012). Teachers' attitudes and students' opposition. School misconduct as a reaction to teachers' diminished effort and affect. Teaching and Teacher Education, 28, 860e869. http://dx.doi.org/10.1016/ j.tate.2012.03.008. De Witte, K., Cabus, S., Thyssen, G., Groot, W., & Maassen van den Brink, H. (2013). A critical review of the literature on school dropout. Educational Research Review, 10, 13e28. http://dx.doi.org/10.1016/j.edurev.2013.05.002. Duquet, N., Glorieux, I., Laurijssen, I., & Van Dorsselaer, Y. (2006). Wit krijt schrijft beter. Schoolloopbanen van allochtone jongeren in beeld [White chalk writes better. Educational careers of allochthonous youngsters in the picture]. Antwerp/Apeldoorn: Garant. Duru-Bellat, M., & Mingat, A. (1997). La constitution de classes de niveau dans les ges: Les effets pervers d'une pratique a  vise
e e
galisatrice [Multilevel colle classrooms in French secondary schools: The perverse effects of a supposedly egalitarian practice]. Revue Française de Sociologie, 38, 759e789. http:// dx.doi.org/10.2307/3322627. Erikson, R., Goldthorpe, J. H., & Portocarero, L. (1979). Intergenerational class mobility in three Western European societies: England, France and Sweden. British Journal of Sociology, 30, 415e441. http://dx.doi.org/10.2307/589632. Evertson, C. M. (1982). Differences in instructional activities in higher- and lowerachieving junior high English and math classes. Elementary School Journal, 82, 329e350. http://dx.doi.org/10.1086/461271. Fall, A. M., & Roberts, G. (2012). High school dropouts: interactions between social context, self-perceptions, school engagement, and student dropout. Journal of Adolescence, 35, 787e798. http://dx.doi.org/10.1016/j.adolescence.2011.11.004. Fang, Z. (1996). A review of research on teacher beliefs and practices. Educational Research, 38, 47e65. Feiman-Nemser, S., & Floden, R. E. (1986). The cultures of teaching. In M. C. Wittrock (Ed.), Handbook of research on teaching (3rd ed., pp. 505e526). New York: Macmillan.
gre
gation ethnique au colle
ge et ses conse
quences [Ethnic Felouzis, G. (2003). La se segregation at secondary school and its consequences]. Revue Française de Sociologie, 44, 413e447. http://dx.doi.org/10.2307/3323202. Finn, J. D. (1972). Expectations and the educational environment. Review of Educational Research, 42, 387e410. Finn, J. D. (1989). Withdrawing from school. Review of Educational Research, 59,

117e142. http://dx.doi.org/10.3102/00346543059002117. Friedkin, N. E., & Thomas, S. L. (1997). Social positions in schooling. Sociology of Education, 70, 239e255. http://dx.doi.org/10.2307/2673266. Frost, M. B. (2007). Texas students' college expectations: does high school racial composition matter? Sociology of Education, 80, 43e66. Gamoran, A. (1993). Alternative uses of ability grouping in secondary schools: can we bring high-quality instruction to low-ability classes. American Journal of Education, 102, 1e22. http://dx.doi.org/10.1086/444056. Gamoran, A., Nystrand, M., Berends, M., & LePore, P. C. (1995). An organizational analysis of the effects of ability grouping. American Educational Research Journal, 32, 687e715. Glick, W. H. (1985). Conceptualizing and measuring organizational and psychological climate: pitfalls in multilevel research. Academy of Management Review, 10, 601e616. http://dx.doi.org/10.2307/258140. Goodenow, C. (1993). The psychological sense of school membership among adolescents: scale development and educational correlates. Psychology in the Schools, 30, 79e90. http://dx.doi.org/10.1002/1520-6807(199301)30:1<79:: AID-PITS2310300113>3.0.CO;2-X. Goodlad, J. I. (1984). A place called school: Prospects for the future. New York: McGraw-Hill. Hallam, S., & Ireson, J. (2006). Secondary school pupils' preferences for different types of structured grouping practices. British Educational Research Journal, 32, 583e599. http://dx.doi.org/10.1080/01411920600775274. Hallinan, M. T. (2008). Teacher influences on students' attachment to school. Sociology of Education, 81, 271e283. Hallinan, M. T., & Kubitschek, W. N. (1999). Curriculum differentiation and high school achievement. Social Psychology of Education, 3, 41e62. Hargreaves, D. H. (1967). Social relations in a secondary school. London: Routledge and Kegan Paul. Hargreaves, A. (1992). Cultures of teaching: a focus for change. In A. Hargreaves, & M. Fullan (Eds.), Understanding teacher development (pp. 216e240). New York: Teachers College Press. Hawker, D. S. J., & Boulton, M. J. (2000). Twenty years' research on peer victimization and psychosocial maladjustment: a meta-analytic review of crosssectional studies. Journal of Child Psychology and Psychiatry and Allied Disciplines, 41, 441e455. http://dx.doi.org/10.1017/S0021963099005545.
bert, T. P., & Reis, S. M. (1999). Culturally diverse high-achieving students in an He urban high school. Urban Education, 34, 428e457. http://dx.doi.org/10.1177/ 0042085999344002. Hinnant, J. B., O'Brien, M., & Ghazarian, S. R. (2009). The longitudinal relations of teacher expectations to achievement in the early school years. Journal of Educational Psychology, 101, 662e670. http://dx.doi.org/10.1037/a0014306. Hinojosa, M. S. (2008). Black-white differences in school suspension: effect of student beliefs about teachers. Sociological Spectrum, 28, 175e193. http:// dx.doi.org/10.1080/02732170701796429. Hofstede, G., Neuijen, B., Ohavy, D. D., & Sanders, G. (1990). Measuring organizational cultures: a qualitative and quantitative study across 20 cases. Administrative Science Quarterly, 35, 286e316. http://dx.doi.org/10.2307/2393392. Hughes, J. N., Gleason, K. A., & Zhang, D. A. (2005). Relationship influences on teachers' perceptions of academic competence in academically at-risk minority and majority first grade students. Journal of School Psychology, 43, 303e320. http://dx.doi.org/10.1016/j.jsp.2005.07.001. Huisman, M. (2000). Imputation of missing item responses: some simple techniques. Quality and Quantity, 34, 331e351. Ireson, J., & Hallam, S. (1999). Raising standards: is ability grouping the answer? Oxford Review of Education, 25(3), 343e358. Janosz, M., Archambault, I., Morizot, J., & Pagani, L. S. (2008). School engagement trajectories and their differential predictive relations to dropout. Journal of Social Issues, 64, 21e40. http://dx.doi.org/10.1111/j.1540-4560.2008.00546.x. Jenkins, P. H. (1995). School delinquency and school commitment. Sociology of Education, 68, 221e239. http://dx.doi.org/10.2307/2112686. Jones, M. G. (1989). Gender issues in teacher education. Journal of Teacher Education, 40, 33e38. http://dx.doi.org/10.1177/002248718904000108. Jussim, L. (1986). Self-fulfilling prophecies. A theoretical and integrative review. Psychological Review, 93, 429e445. http://dx.doi.org/10.1037//0033295X.93.4.429. Jussim, L., & Harber, K. D. (2005). Teacher expectations and self-fulfilling prophecies: knowns and unknowns, resolved and unresolved controversies. Personality and Social Psychology Review, 9, 131e155. http://dx.doi.org/10.1207/ s15327957pspr0902_3. Kelly, S. (2008). What types of students' effort are rewarded with high marks? Sociology of Education, 81, 32e52. Kelly, S., & Carbonaro, W. (2012). Curriculum tracking and teacher expectations: evidence from discrepant course taking models. Social Psychology of Education, 15, 271e294. http://dx.doi.org/10.1007/s11218-012-9182-6. Kornblau, B. (1982). The teachable pupil survey: a technique for assessing teachers' perceptions of pupil attributes. Psychology in the Schools, 19, 170e174. http:// dx.doi.org/10.1002/1520-6807(198204)19:2<170::AIDPITS2310190206>3.0.CO;2-P. Lamb, S., Markussen, E., Teese, R., Sandberg, N., & Polesel, J. (Eds.). (2011). School dropout and completion. International comparative studies in theory and policy. Dordrecht/New York: Springer. Lamote, C., Van Damme, J., Van Den Noortgate, W., Speybroeck, S., Boonen, T., & de Bilde, J. (2013). Dropout in secondary education: an application of a multilevel discrete-time hazard model accounting for school changes. Quality and

M. Van Houtte, J. Demanet / Teaching and Teacher Education 54 (2016) 117e127 Quantity, 47, 2425e2446. Lee, V. E., & Burkam, D. T. (2003). Dropping out of high school: the role of school organization and structure. American Educational Research Journal, 40, 353e393. http://dx.doi.org/10.3102/00028312040002353. Linchevski, L., & Kutscher, B. (1998). Tell me with whom you're learning, and I'll tell you how much you’ve learned: mixed-ability versus same-ability grouping in mathematics. Journal for Research in Mathematics Education, 29, 533e554. http://dx.doi.org/10.2307/749732. Luyten, H., Bosker, R., Dekkers, H., & Derks, A. (2003). Dropout in the lower tracks of Dutch secondary education: predictor variables and variation among schools. School Effectiveness and School Improvement, 14, 373e411. http://dx.doi.org/ 10.1076/sesi.14.4.373.17158. Malmberg, L. E., & Trempala, J. (1997). Anticipated transition to adulthood: the effect of educational track, gender, and self-evaluation on Finnish and Polish adolescents' future orientation. Journal of Youth and Adolescence, 26, 517e537. http://dx.doi.org/10.1023/A:1024577805149. McLaughlin, M. W. (1993). What matters most in teachers' workplace context? In J. W. Little, & M. W. McLaughlin (Eds.), Teachers' work: Individuals, colleagues, and contexts (pp. 79e103). New York: Teachers College Press. McNeal, R. B., Jr. (1997). High school dropouts: a closer examination of school effects. Social Science Quarterly, 78, 209e222. Metz, M. H. (1993). Teachers' ultimate dependence on their students. In J. W. Little, & M. W. McLaughlin (Eds.), Teachers' work: Individuals, colleagues, and contexts (pp. 104e136). New York: Teachers College Press. Midgley, C., Feldlaufer, H., & Eccles, J. S. (1988). The transition to junior high school: beliefs of pre- and post-transition teachers. Journal of Youth and Adolescence, 17, 543e562. http://dx.doi.org/10.1007/BF01537831. Natriello, G., Pallas, A. M., & Alexander, K. L. (1989). On the right track? curriculum and academic achievement. Sociology of Education, 62, 109e118. http:// dx.doi.org/10.2307/2112843. Nixon, D. (2006). ‘I just like to work with my hands’: employment aspirations and the meaning of work for low-skilled unemployed men in Britain's service economy. Journal of Education and Work, 19, 201e217. Oakes, J. (1985). Keeping track: How schools structure inequality. New Haven, CT: Yale University Press. Oakes, J. (2005). Keeping track: How schools structure inequality (2nd ed.). New Haven, CT: Yale University Press. Page, R. (1991). Lower track classrooms: A curricular and cultural perspective. New York: Teachers College Press. Rosenbaum, J. E. (1976). Making inequality: The hidden curriculum of high school tracking. New York: Wiley. Rosenbaum, J. E. (2001). Beyond college for all: Career paths for the forgotten half. New York: Russell Sage Foundation. Rosenthal, R. (2002). The Pygmalion effect and its mediating mechanisms. In J. M. Aronson (Ed.), Improving academic achievement: Impact of psychological factors on education (pp. 25e36). London: Academic Press. Rosenthal, R., & Jacobson, L. (1968). Pygmalion in the classroom. New York: Holt, Rinehart & Winston. Rubie-Davies, C., Hattie, J., & Hamilton, R. (2006). Expecting the best for students: teacher expectations and academic outcomes. British Journal of Educational Psychology, 76, 429e444. http://dx.doi.org/10.1348/000709905X53589. Rubovits, P. C., & Maehr, M. L. (1971). Pygmalion analyzed. Toward an explanation of Rosenthal-Jacobson findings. Journal of Personality and Social Psychology, 19, 197e203. http://dx.doi.org/10.1037/h0031526. Rumberger, R. W. (1987). High school dropouts: a review of issues and evidence. Review of Educational Research, 57, 101e121. http://dx.doi.org/10.3102/ 00346543057002101. Rumberger, R. W. (1995). Dropping out of middle school: a multilevel analysis of students and schools. American Educational Research Journal, 32, 583e625. http://dx.doi.org/10.3102/00028312032003583. Rumberger, R. W., & Palardy, G. J. (2005). Test scores, dropout rates, and transfer rates as alternative indicators of high school performance. American Educational Research Journal, 42, 3e42. http://dx.doi.org/10.3102/00028312042001003. Rumberger, R. W., & Thomas, S. L. (2000). The distribution of dropout and turnover rates among urban and suburban high schools. Sociology of Education, 73, 39e67. http://dx.doi.org/10.2307/2673198. Schafer, W., & Olexa, C. (1971). Tracking and opportunity. The locking-out process and beyond. Scranton, PA: Chandler. Shavit, Y., & Featherman, D. L. (1988). Schooling, tracking, and teenage intelligence. Sociology of Education, 61, 42e51. http://dx.doi.org/10.2307/2112308. Shrout, P. E., & Fleiss, J. L. (1979). Intraclass correlations: uses in assessing rater reliability. Psychological Bulletin, 86, 420e428. http://dx.doi.org/10.1037//00332909.86.2.420. [Education on the way in migration society]. In Sierens, S., Van Houtte, M., Pelleriaux, K., Loobuyck, P., & Delrue, K. (Eds.), Onderwijs onderweg in de immigratiesamenleving. Gent: Academia Press.

127

Stearns, E., & Glennie, E. J. (2006). When and why dropouts leave high school. Youth & Society, 38, 29e57. http://dx.doi.org/10.1177/0044118X0528764. Stevens, P. A. J. (2007). Exploring the importance of teachers' institutional structure on the development of teachers' standards of assessment in Belgium. Sociology of Education, 80, 314e329. Stevens, P. A. J., & Vermeersch, H. (2010). Streaming in Flemish secondary schools: exploring teachers' perceptions of and adaptations to students in different streams. Oxford Review of Education, 36, 267e284. http://dx.doi.org/10.1080/ 03054981003629862. Tas, A., Selvitopu, A., Bora, V., & Demirkaya, Y. (2013). Reasons for dropout for vocational high school students. Educational Sciences. Theory & Practice, 13, 1561e1565. Timmerman, C., Hermans, P., & Hoornaert, J. (2002). Allochtone jongeren in het onderwijs. Een multidisciplinair perspectief [Allochthonous youngsters in education: A multidisciplinary perspective]. Leuven-Apeldoorn, The Netherlands: Garant. €ller, O., & Baumert, J. (2006). Tracking, Trautwein, U., Lüdtke, O., Marsh, H. W., Ko grading, and student motivation: using group composition and status to predict self-concept and interest in ninth-grade mathematics. Journal of Educational Psychology, 98, 788e806. http://dx.doi.org/10.1037/0022-0663.98.4.788. Trouilloud, D. O., Sarrazin, P. G., Martinek, T. J., & Guillet, E. (2002). The influence of teacher expectations on student achievement in physical education classes: Pygmalion revisited. European Journal of Social Psychology, 32, 591e607. http:// dx.doi.org/10.1002/ejsp.109. van Uden, J. M., Ritzen, H., & Pieters, J. M. (2014). Engaging students: the role of teachers' beliefs and interpersonal teacher behavior in fostering student engagement in vocational education. Teaching and Teacher Education, 37, 21e32. http://dx.doi.org/10.1016/j.tate.2013.08.005. Van Houtte, M. (2004). Tracking effects on school achievement: a quantitative explanation in terms of the academic culture of school staff. American Journal of Education, 110, 354e388. http://dx.doi.org/10.1086/422790. Van Houtte, M. (2006a). School type and academic culture: quantitative evidence for the differentiation-polarisation theory. Journal of Curriculum Studies, 38, 273e292. http://dx.doi.org/10.1080/00220270500363661. Van Houtte, M. (2006b). Tracking and teacher satisfaction: the role of study culture and trust. Journal of Educational Research, 99, 247e254. Van Houtte, M. (2011). So where’s the teacher in school effects research? The impact of teachers’ beliefs, culture and behaviour on equity and excellence in education. In K. Van den Branden, P. Van Avermaet, & M. Van Houtte (Eds.), Equity and excellence in education. Towards maximal learning opportunities for all students (pp. 75e95). , New York: Routledge. Van Houtte, M., Demanet, J., & Stevens, P. A. J. (2013). Curriculum tracking and teacher evaluations of individual students: selection, adjustment or labeling? Social Psychology of Education, 16, 329e352. http://dx.doi.org/10.1007/s11218013-9216-8. Van Houtte, M., & Stevens, P. A. J. (2008). Sense of futility: The missing link between track position and self-reported school misconduct. Youth & Society, 40, 245e264. http://dx.doi.org/10.1177/0044118X08316251. Van Houtte, M., & Stevens, P. A. J. (2009). Study involvement of academic and vocational students: does between-school tracking sharpen the difference? American Educational Research Journal, 46, 943e973. http://dx.doi.org/10.3102/ 0002831209348789. Van Houtte, M., & Stevens, P. A. J. (2010). The culture of futility and its impact on study culture in technical/vocational schools in Belgium. Oxford Review of Education, 36, 23e43. http://dx.doi.org/10.1080/03054980903481564. Van Houtte, M., & Stevens, P. A. J. (2015). Tracking and sense of futility: the impact of between school tracking versus within school tracking in secondary education in Flanders (Belgium). British Educational Research Journal, 41, 782e800. http:// dx.doi.org/10.1002/berj.3172. Van Houtte, M., & Van Maele, D. (2012). Students' sense of belonging in technical/ vocational schools versus academic schools: the mediating role of faculty trust in students. Teachers College Record, 114, 1e36. Van Landeghem, G., De Fraine, B., Gielen, S., & Van Damme, J. (2013). Vroege schoolverlaters in Vlaanderen in 2010. Indeling volgens locatie, opleidingsniveau van de moeder en moedertaal [Early school leavers in Flanders in 2010. Classification according to location, mother's education and mother tongue]. Leuven: Steunpunt SSL. rapport nr. SSL/2013.05/1.2.0. Van Landeghem, G., & Van Damme, J. (2011). Vroege schoolverlaters uit het voltijds beroepsonderwijs. Vertrekpunt voor een gedifferentieerde aanpak [Early school leavers in full-time vocational education. Starting point for a differentiated approach]. Leuven: Steunpunt SSL. Wehlage, G. G., & Rutter, R. A. (1986). Dropping out: how much do schools contribute to the problem? Teachers College Record, 87, 374e392. Worrell, F. C., & Hale, R. L. (2001). The relationship of hope in the future and perceived school climate to school completion. School Psychology Quarterly, 16, 370e388.